ROA Technical Report. International mobility of students. Didier Fouarge Merve Özer

Research Centre for Education and the Labour Market ROA International mobility of students Its impact on labour market forecasts and its contribution to the Dutch economy Didier Fouarge Merve Özer ROA Technical Report ROA-TR-2014/6 Researchcentrum voor Onderwijs en Arbeidsmarkt ROA Research Centre for Education and the Labour Market ROA

International mobility of students Its impact on labour market forecasts and its contribution to the Dutch economy Didier Fouarge Merve Özer ROA-TR-2014/6 December 2014 Research Centre for Education and the Labour Market Maastricht University P.O. Box 616, 6200 MD Maastricht, The Netherlands T +31 43 3883647 F +31 43 3884914 secretary-roa-sbe@maastrichtuniversity.nl www.roa.nl

Table of Contents Page Table of Contents 1 1 Introduction 2 2 Labour market impacts of international students: A literature review 5 2.1 Types of study migration 5 2.2 Benefits of degree mobility in terms of labour market: A review on retention of degree-mobile students 7 2.3 Potential costs of degree mobility in terms of labour market 10 3 Student mobility in the Netherlands 12 3.1 Incoming student mobility 12 3.2 Outgoing student mobility 19 3.3 Net study migration 23 4 Sensitivity of labour markets forecasts to study migration 24 4.1 Inflow of graduates and degree migration 24 4.2 Sensitivity of forecasts 26 5 Conclusion 30 References 31 Appendix: Further dimensions of study mobility 35 A.1 Costs of student mobility 36 A.2 Benefits of student mobility 37 1

1 Introduction International student mobility has rapidly increased in the past three decades: the number of students enrolled in tertiary education outside of their country of citizenship was 0.8 million in 1975, but it increased to 4.3 million in 2011 (OECD, 2013). 1 This rapid increase can be explained by several factors: internationalization and standardization of higher education (e.g., the Bologna process in European Union countries), global increase in demand for tertiary education, faster information flows thanks to the advancements in communication and transportation opportunities, and policies to encourage student exchange through bilateral agreements (Tremblay, 2005; OECD, 2013; Van Bouwel & Veugelers, 2013). For students, there are several benefits to engaging in higher education abroad. International study mobility is an opportunity to achieve education in fields that are unavailable in the home country or are known to be of better quality in another country, to increase their future employment opportunities and earnings, or to improve their language and intercultural skills (Fouarge and van Thor, 2011; Varghese, 2008). While study mobility is an investment decision made by students for their own self-development and future career opportunities, it has important implications in terms of sending and receiving countries in a broader context of human capital accumulation of a country. From the sending countries perspective, student mobility might be used as a means of transferring new knowledge and technology to their countries if migrant students return back to their home countries or establish networks with their country of origin (Tremblay, 2005; Gribble, 2008). Also the host countries might benefit from student mobility because of expanded available resources for their higher education institutions, enhanced productivity of the labour market inflow of student due to positive selection in study migration, improved international reputation of their higher education systems, or because those students can be employed in the host country after graduation, and student mobility can be positioned in a broader strategy of attracting high skilled expatriates (Tremblay, 2005; Centraal Planbureau, 2012; Abella, 2006). As industrialized economies in the context of globalization have been evolving towards knowledge-based economies, having a sufficiently large stock of skilled manpower has become a key concern to keep pace with competitors. This has important implications in terms of brain migration considering skill shortages faced by those countries especially in the context of an aging population. Being aware of this fact, many OECD countries have put several measures into effect in order to relax their migration regulations in favour of sustaining temporary or permanent stay of high skilled migrants since the late 1990s (see Tremblay, 2005; Abella, 2006; Kahanec & Zimmermann, 2011). Student mobility is an important complement of these efforts considering the large and increasing pool of foreign students in several OECD countries, of which some are currently pursuing active policies to attract high 1 This report was written as part of the Project Onderwijs-Arbeidsmarkt (POA) of ROA. We thank the members of the POA advisory board, especially Djoerd de Graaf, for the useful comments provided on an earlier draft. 2

skilled migrants (OECD, 2013). Considering that migration is a selective process (Aslanbeigui & Montecinos, 1998; Grogger & Hanson, 2013), incoming students provide the opportunity to the host countries to attract the best students among many good students as potential future high skilled migrants. This means that student mobility is of strategic importance as a policy for attracting highly skilled expatriates. Although the flow of international students to developed OECD countries has been rising, a portion of those students prefer to return back to their home countries or to migrate to another country. This leads to discussions in the literature regarding the validity of concepts such as brain drain, brain gain, or brain circulation (see Straubhaar, 2000; Rizvi, 2005; Lee & Kim, 2010; Gibson & McKenzie, 2011). Even if the proportion of international students intending to stay in their host countries is initially high, actual stay rates might not coincide with those rates (Sykes & Chaoimh, 2012). This is because the decision to stay, to return, or to migrate another country is affected by relative opportunities in labour markets of different economies as well as family, cultural, and political factors (Lee & Kim, 2010; Bratsberg, 1995). In this respect, how and to what extent international students contribute to the national economy and labour market after their graduation are important questions from the perspective of host countries. Considering the labour market shortages and implementation of policies aiming to fill those shortages in OECD countries, the retention rate of international students after graduation and their incorporation in the labour market become key issues that determine the extent to which countries receiving large flows of international students are able to make use of this opportunity to attract them as highly-skilled immigrants. The Netherlands is one of those countries that face an aging labour force and skill shortages in particular sectors. Therefore, the Netherlands is currently pursuing policies to ease migration procedures of high-skilled immigrants as well as to attract more international students. As a result of the Bologna process and the promotion activities in some non-eu countries, the number of international students has exhibited an increasing trend in recent years just like in other countries around the world (Becker & Kolster, 2012). In addition, a new scheme came into force in 2007 according study migrants the right to stay for one more year in the Netherlands in order to search for a suitable job. The aim is that international graduates from Dutch bachelor s and master s programs obtain the highly-skilled migrant status (Sykes & Chaoimh, 2012). So besides the efforts to attract international students, the Netherlands is increasing its efforts to keep those students in productive employment after graduation. In this regard, the retention rate of international students and their transition to productive employment is a key issue also for Dutch economy. In this study, we first discuss the contribution of study migrants to the labour market. This contribution crucially depend on the retention rate of study migrants in the Dutch labour market. Then we investigate the impact study migration has on ROA s labour market forecasts by educational type. The number of international students staying in the Netherlands after graduation is crucial when making labour market forecasts for the inflow into the labour market: if foreign students completing their degree are counted as inflow as is the case in the Referentieramingen (Ministerie van 3

Onderwijs, Cultuur en Wetenschap, 2013) and the retention rate is less than one, then one would overestimate the true labour supply. To assess the sensitivity of ROA s labour market forecasts, we take the inflows of international graduates to the Dutch labour market into account and estimate the impact of those inflows on the balance between supply and demand balance in the following six years by field of education. The rest of the paper is organized as follows: Section 2 presents a literature review on the potential impacts of international graduates on labour market, with a specific focus on the retention of degree-mobile students in the Netherlands. Recent data and trend information on student mobility in the Netherlands is discussed in Section 3. In Section 4, the sensitivity of ROA s labour market forecasts by fields of education to the inclusion of foreign students who take their degree in the Netherlands is discussed. Section 5 concludes. 4

2 Labour market impacts of international students: A literature review 2.1 Types of study migration The labour market contribution of international mobile students crucially depend on the extent to which they stay in their host country upon graduation. Studies on retention rates of international students and proportion of foreign students intending to stay in the host country provide a variety of results that are difficult to compare and inconclusive to derive a clear measure of retention. This makes evaluating potential labour market contribution of international students difficult. One of the reasons is that the data sources used cover different groups of international students by type of mobility and/or level of study. Therefore, classifying internationally mobile students by the type of mobility (and also level of study) is important before discussing the potential contribution of student mobility to the host country s labour market. This is because intention of staying and actual retention might exhibit significant differences between certain groups of international students. In this report, international students are grouped under two broad categories: credit-mobile and degree-mobile students. Credit-mobile students are mainly exchange students who are enrolled in higher education institutions and go abroad to study in a host institution for a limited period of time to earn credits under some bilateral agreements or exchange program schemes. The most well-known exchange program scheme is the Erasmus Program of the European Commission that provides the opportunity to students to go abroad for a period of 3 to 12 months at any stage of their study cycle. Credit-mobile students are mostly concentrated in bachelor s degree programs (European Commission, 2013). They enrol in programs or courses in host institutions for a limited period of time, receive a scholarship to cover their travel and subsistence costs, and are exempted from any kind of fees for their studies (European Commission, 2014). The second group of international students is degree-mobile students who enrol in a full-time bachelor s and/or master s program of study with the aim to obtain a diploma at the end of their study. The literature sometimes also considers PhD students to be degree-mobile students. Although all of those students are longer-term migrants compared to the limited duration of credit-mobile students in the host country, PhD students differ from degree-mobile students in bachelor s and master s programs. While bachelor s and master s students are regular students who pay their tuition fees to enrol in study programs, PhD students are usually employed by their host institution and involved in productive activities in the form of research and teaching. The degree of labour market contribution of study migration is expected to differ by groups of international students as mentioned above. First of all, credit-mobile students are not expected to participate in the labour market in the host country due to the following reasons. As mentioned earlier, most of the credit-mobile students are 5

in bachelor s degree programs (European Commission, 2013), their duration of stay in the host country is usually short, and they eventually return back to their home countries since they are already enrolled in study programs in their home countries. However, it should be noted that students spending some of their study time abroad through exchange programs are more likely to migrate again after their graduation (de Grip et al., 2010; Parey & Waldinger, 2011). So if credit-mobile students prefer to migrate to their earlier host countries after graduation (due to the familiarity they gain during their exchange visits), credit mobility might be a precursor of further degree or high-skilled labour mobility. Secondly, although PhD students are sometimes considered as mobile students in the literature as mentioned above, they are actually high skilled migrants considering their participation in productive activities in the labour market for higher education. International PhD students coming to the Netherlands have been benefiting from regulations for skilled migrants since 2004 (Sykes & Chaoimh, 2012) so they are a part of labour market just like skilled immigrant workers. Although the retention of PhD students after graduation is important for host countries being in need of skilled immigrants to fill labour market shortages, it could be analysed as a long-term stay decision of skilled migrants rather than retention of mobile students. Lastly, degree-mobile bachelor s and master s students might participate into the host country s labour market while studying through internship programs as a part of their study program. However, it should be noted that those internship programs are not required in all study programs so many incoming students fall out of this category (Tremblay, 2005). Although non-eu students are allowed to work only for a limited number of hours per week if they have a separate work permit (Becker & Kolster, 2012), EU students do not face such restriction. However, degree-mobile bachelor s and master s students are not expected to participate extensively into the labour market and do productive work since they mostly attend full-time education. This means that the potential contribution of degree-mobile students to the host country s labour market crucially depends on their retention after graduation. In light of the above distinction in types of study mobility, this report concentrates on the labour market contribution of degree mobility, and particularly on the retention of international students in bachelor s and master s degree programs. In this respect, international bachelor s and master s students constitute a specific group in terms of ROA s labour market forecasts since this group s transition to the Dutch labour market has important implications for gains and losses of the Netherlands from student mobility as a host country. Also, the inflow of degree mobile students on the Dutch labour market could potentially affect the labour market prospects of nationals (see Section 4). 6

2.2 Labour market benefits of degree mobility: retention rates of degreemobile students Student mobility might be a channel for attracting future high-skilled migrants to the labour market. 2 Abella (2006) refers to this strategy as academic-gate approach in the sense that students who come to a host country for their study are highly qualified due to the selective nature of migration process and they can enter to the labour market at a relatively low cost. This might be due to the fact that obtaining a diploma from a university in the host country increases the chance that their skills be acknowledged by institutions offering post-graduate degrees and companies in the host country (Tremblay, 2005). Thus, student mobility might be also used by students as a means of entering into the labour market of a host country (Tremblay, 2005). International evidence on the retention of study migrants The US, as a country with a long immigration tradition, seems benefiting mostly from retention of international graduates. According to an earlier study conducted by U.S. Immigration and Naturalization Service in 2000, 23% of current temporary work permission (H-1B) holders at that time was previously holding a student visa. Another survey conducted by Barucha et al. (2007) among master s students in US and UK universities also showed that only 30.5% of foreign students are planning to return to their home countries right after completing their study program, but that the intention to stay in the US is higher compared to that in UK. Retention of PhD graduates is also very high in the US: 68% of PhD earners in 2006 were still in the US in 2011 (Finn, 2014). 10-year stay rate calculated for the 2001 PhD earners cohort is 65%, but it differs by filed of study: the stay rate is highest among graduates from engineering fields (Finn, 2014). Finn (2014) showed that retention rate for temporary visa holders is similar to those figures and intention to stay is a good predictor of the actual retention rate among PhD graduates in the US. Intentions of international students to stay in Western European countries seem also high, at least in the short term, but lower than in the US. A survey conducted in 2011 among non-eu master s and PhD students in five EU countries showed that most of the students have intentions to stay at least for 1-2 years after completing their studies mainly to gain international work experience (Sykes & Chaoimh, 2012), which provides evidence that student mobility constitutes a path for entry into the labour market. According to the results of this survey, 35.2% and 20.1% of international master s and PhD students in the Netherlands are planning to stay for 1-2 and 3-5 years more, respectively. Only 6.1% of them are planning to stay more than 5 years (Sykes & Chaoimh, 2012). Just like in the US, this study shows that students in technical fields are more inclined to stay (Sykes & Chaoimh, 2012). This is to be expected considering skill shortages in technical sectors and the related promising labour market opportunities. However, to the best of our knowledge, there is lack of information regarding the actual stay rates of those post-graduate students in Europe 2 In an appendix, we discuss other costs and benefits of student mobility for the host country. 7

so that a comparison with intentions as well as with retention rate in the US is not possible. Intentions of stay and actual retention rates might differ by destination country and country of origin as well as level an field of study. Data on several OECD countries reveal that short-term retention rates of degree-mobile students (at bachelor s and master s level) are below 35% as presented in Figure 2.1. This indicator provided by OECD (2011) was calculated as the proportion of international students changing their visa status from student visa to students who did not renew their visas in 2008 (or in 2009 for some countries). Although this indicator is not fully reliable because of inconsistencies between data sources, it provides an insight for short-term retention trends in several OECD countries (OECD, 2011). According to Figure 2.1, the retention rate in the Netherlands is above 25% which is the OECD average, and the Netherlands is one of the countries that achieves the highest rentention rate. In addition, it was estimated that 80% or more of those changes in visa status in the Netherlands are work-related (OECD, 2011), which suggests that the retention of foreign students benefits the Dutch labour market. Figure 2.1: Percentage of foreign students changing visa status in 2008-2009 Source: OECD, 2011 Retention rates of study migrants in the Netherlands Besides OECD s short-term retention rate based on simple accounting, there are several other studies addressing the stay rates of international students in the Netherlands. These studies make use of administrative records and/or survey data. One of those studies was conducted by Bijwaard (2010) by using register data for the Netherlands for the period 1995-2003. The author estimated that the average duration of the stay of a migrant student in the Netherlands after graduation is 29 months. 45% of study migrants left the Netherlands during the observation period while in the long term, only 19% of students permanently stay in the Netherlands. According to Dutch regulation, every person who enters to the country and stays at 8

least 4 months should register with the local municipality. Thus, the register data used by Bijwaard (2010) cover all migrants who stay at least 4 months in the Netherlands, thereby including credit-mobile students (e.g., under the Erasmus scheme) who are supposed to stay in the Netherlands from 3 to 12 months. In this respect, expected short period of stay for credit-mobile students might cause a downward bias in the estimates of Bijwaard (2010). The estimated 19% permanent retention rate might therefore be considered as a lower bound. Other estimates available for incoming students in the Netherlands give much higher retention rates compared to that of Bijwaard (2010). A summary of the findings of those studies is presented in a Dutch report prepared by Centraal Planbureau (2012). According to the results of HBO and WO Monitor conducted in 2007-2011, the retention rate 1,5 years after graduation is 43% while micro data analysis conducted by CBS in 2010 shows that 38% and 34% of foreign graduates remained and were employed in the Netherlands after 3-4 and 13-15 years of graduation, respectively. Results from the UM-Scanner conducted among Maastricht University graduates in 2010 show that the retention rate of EU students is 22% and 29% in 5-7 and 9-12 years after graduation, respectively. These rates might give an idea about the retention rates of degree-mobile students in the short-, medium-, and long-term but should be taken with caution since these calculations might suffer from selection bias (see Centraal Planbureau, 2012). In brief, estimates on retention rate of degree-mobile international students in the Netherlands range from 19% to 43% which might be considered as lower and upper bounds for the retention rate. As mentioned earlier, 19% rate of retention estimated by Bijwaard (2010) is based on the data covering credit-mobile students who are inclined to leave the host country after a limited period time so this might be a downward-biased estimate. However, the 43% rate of retention estimated from HBO and WO Monitor reflects the retention only 1,5 years after graduation. Since intention to stay and actual retention might be high in early years following graduation considering the Dutch regulation allowing international students to stay one more year after graduation, this rate might represent the potential maximum of retention in the Netherlands. In Section 4, we take these retention rates into account to test the sensibility of ROA s labour market forecast to various assumptions concerning the retention of study migrants in the Dutch labour market. However, it should be noted that the estimates mentioned above do not provide the opportunity of accurately evaluate whether estimated retention of international students are sufficient to alleviate labour market shortages in the Dutch economy. This is because of issues related to coverage of sample and selectivity, and lack of detail concerning the retention by field of study. In this respect, estimating the level of retention by level and field of study, and by type of mobility through the use of a representative sample or administrative records in the Netherlands might be an important topic for future research to better evaluate the impacts of international students on the Dutch labour market. It is further unknown to what extent the retention of international students varies over the business cycle. This could also be subject of future research. 9

2.3 Potential labour market costs of degree mobility As discussed so far, attracting highly skilled migrants is perceived as a solution to alleviate the pressure of aging population on labour force and skill shortages in labour market in many OECD countries, and student mobility might be a supporting mechanism for these attempts if retention rates are high enough to fill labour market shortages. In this respect, it is argued in the previous section that the labour market might benefit from international student mobility depending on the level of retention of international students. Nevertheless, it could be claimed that retention of international students after graduation, and their transition to labour market might negatively affect the employment opportunities of native graduates. This could be considered as a cost associated with degree mobility in terms of host country s labour market. This view is supported by Borjas (2004) who argues that enrolment of foreign PhD students in the US causes a displacement of white native American males. However, there is no consensus in the literature in the sense that studies conducted so far on so-called displacement effects do not provide any definite evidence on the negative effects of immigrants on natives labour market outcomes (see Kerr & Kerr, 2011). In the context of the Dutch labour market, skill shortages are especially expected in specific segments of the technical and health care sector (ROA, 2013). As mentioned earlier, a survey conducted among master s and PhD students in five EU countries reveal that students in technical fields are more inclined to stay (Sykes & Chaoimh, 2012) apparently due to promising employment prospects in those sectors. In case of skill shortages a potential displacement of native graduates due to retention of international students seems unlikely as these graduates will fill the gap between the labour demand by firms and the labour supply of nationals. Even if it is assumed that there is a displacement effect, it should be considered with caution because there are other potential positive impacts of international student mobility, for example in terms of employment creation in the host country. According to a NAFSA report (2013), one international student enrolled in a US college or university created 0.43 jobs in sectors such as higher education, accomodation, and telecommunication through his/her spendings in 2012/13. A similar study by Münch and Hoch (2013) for the Netherlands reveals that average expenditure of an international student for consumption and accommodation created 0.15 annual additional jobs in 2010/11 in the Netherlands. Münch and Hoch (2013) also estimated that 0.29 annual additional jobs can be created through consumption spending per international student entering into the Dutch labour market after graduation. In this regard, the impact of international student retention on the host country s labour market should be evaluated by the relative magnitudes of job created and job displacement of natives. Another cost aspect of degree mobility is that associated with outgoing study migration. While incoming students might constitute a potential gain for their host countries labour markets, outgoing students might be a loss in terms of their home countries workforce. In this respect, the number of outgoing students relative to 10

incoming students and their retention rate in their host countries (or their return rate to their home countries) become key issues for OECD countries experiencing labour market shortages. In the Netherlands, the number of incoming students outweighs the number of outgoing students (see Section 3). However, the case of outgoing degree-mobile students has been neglected in this study because of the lack of data on return rates. 11

3 Student mobility in the Netherlands Each year, the Netherlands is attracting more and more international students from both EU and non-eu countries (Becker & Kolster, 2012). Although tuition fees are higher, especially for non-eu students, compared to some other European countries, the Netherlands has fully adopted the European bachelor s/master s degree structures following the Bologna process, and has many English-taught bachelor s and master s programs (Becker & Kolster, 2012). In order to attract students from developing countries, Nuffic Neso offices in 10 countries such as Brazil, India, Vietnam and Thailand carry out promotion activities, and the Netherlands also tries to attract those non-eu students as high-skilled migrants by allowing them to stay one more year after their graduation to look for a job (Becker & Kolster, 2012). As a result of these efforts, student mobility has increased in recent years. We document these trends in this section. 3.1 Incoming student mobility Trends in degree mobility As Figure 3.1 shows, the number of foreign students in the Netherlands has been increasing since 2005. The main source of this increase is the increasing number of degree-mobile students rather than exchange students coming for study or work placement under the Erasmus program. Almost 80% of all incoming students are degree mobile students who have come to the Netherlands with the aim of obtaining a diploma. 12

Figure 3.1: Number of foreign students in the Netherlands in 2005-2013 Source: Nuffic, 2014 The distribution of degree-mobile students over higher professional education (HBO University of Applied Sciences) and academic higher education (WO Research University) is given in Figure 3.2. In the academic year 2012/13, more than 58,000 foreign students were enrolled in higher education in the Netherlands and they were distributed almost equally between research universities and universities of applied sciences. The figure shows that the share of foreign students in total enrolment at research universities has always been higher than the share of foreign students in universities of applied sciences. From 2008 to 2013, the share of foreign students in total enrolment has increased from 6.2% to 6.9% in HBOs while it has increased from 8.6% to 12.2% in WOs. The share of foreign degree-mobile students in total enrolment in the Netherlands has increased from 7.1% to 8.8% in that period. To put these percentages into perspective, according to UNESCO data, the share of international students in higher education in the US was about 3% in the 2000s, while it increased from 10 to 15% in the UK between 2002 and 2009 (Choudaha and Chang, 2012). In Australia, this share ranged between 17 and 21% in the 2000s. 13

Figure 3.2: Number of degree-mobile foreign students and percentage share in total enrolment in the Netherlands by HBO and WO in 2008-2013 Source: Nuffic, 2014 At research universities (Figure 3.3), the number of degree mobile students in Bachelor s and the Master s programs is more or less similar. However, the increase in degree mobility between 2008 and 2013 in those universities has been higher in Master s programs (Figure 3.3). As Figure 3.4 shows, almost all (95%) degree mobile students in universities of applied sciences in the Netherlands are engaged in Bachelor s programs, and their number has been increasing over the past few years. The number of degree mobile students in HBO master s programs, however, has hardly changed between 2008 and 2013. 14

Figure 3.3: Number of degree-mobile foreign students in WOs by level of study in 2008-2013 Source: Nuffic, 2014 Figure 3.4: Number of degree-mobile foreign students in HBOs by level of study in 2008-2013 Source: Nuffic, 2014 Degree mobility by field of study Foreign students in the Netherlands who enrolled to government-funded WOs are concentrated in several fields of study as reflected in Figure 3.5. Apart from cross- 15

sectoral fields of studies where a third of all students are degree mobile students 3, the share of degree mobile students in 2012/2013 is particularly high in agriculture and natural environment studies (23.6%) and economics (21.2%). The increase in degree mobility in economics between 2008 and 2013 is spectacular: it rose from 15.2% to 21.2%. Figure 3.5: Percentage share of foreign diploma students in WOs by field of study in 2008-2013 Source: Nuffic, 2014 In government-funded HBOs, the share of foreign students is especially large in language and culture. Although it has exhibited a declining trend between 2008 and 2010, the share of degree mobile students in that field equals 25% in 2012/2013 (Figure 3.6). In economics, the share of degree mobile students in universities of applied sciences has increased from 7.7% in 2008 to 9% in 2013. 3 Note that is a relatively small field. 16

Figure 3.6: Percentage share of foreign diploma students in HBOs by field of study in 2008-2013 Source: Nuffic, 2014 Sending countries The Netherlands has been receiving many foreign degree-mobile students from other European Union countries as Figure 3.7 shows. In particular, the Netherlands hosts large numbers of German students in part due to the geographical proximity and in part due to the good reputation of Dutch Universities. However, a large and increasing number of degree-mobile students come from China. Figure 3.7: Top 8 countries of origin of foreign diploma students in the Netherlands in 2008-2013 Source: Nuffic, 2014 17

Incoming credit mobility Although credit mobility is not expected to have a large impact on host countries labour market 4, we now briefly discuss recent trends in incoming credit mobility. Just like the number of degree-mobile students, the number of credit-mobile students has also been increasing since 2005, especially between 2007/08 and 2010/11 (Figure 3.8). The number of incoming credit-mobile students under the Erasmus program increased by almost 20% from 2007/08 to 2012/13. Figure 3.8: Number of inbound Erasmus students in the Netherlands in 2005-2013 Source: Nuffic, 2014 4 The average stay of a degree mobile student is only 6 months (European Commission 2012). 18

Figure 3.9: Percentage of total Erasmus population who prefer the Netherlands for top 5 countries with the highest percentage in 2006-2011 Source: Nuffic, 2014 When looking at absolute numbers, Spain, France, Germany, Italy, and Turkey seem as the main sending countries (Nuffic, 2014). However, considering the share of exchange students in the total Erasmus population in a country who prefer coming to the Netherlands, Sweden, Finland, Belgium, Turkey, and United Kingdom are among the main student-sending countries (Figure 3.9). 3.2 Outgoing student mobility Although outbound student mobility is lower in size compared to inbound student mobility in the Netherlands, an increasing trend in outbound student mobility is visible in recent years. As shown in Figure 3.10, outbound student mobility in the Netherlands has increased by 47% between 2005 and 2011. This increase has been driven by both degree- and credit-mobile students while increase in outbound credit mobility is slightly higher in the same period (almost 50%). 19

Figure 3.10: Number of outgoing students in the Netherlands in 2005-2011 Source: Nuffic, 2014 Outgoing degree mobility Between 2006 and 2011, the number of degree-mobile Dutch students increased by 40% to almost 21,000 as reflected by Figure 3.10. Outgoing Dutch degree-mobile students constituted slightly more than 3% of total enrolment in Dutch universities after 2009. 20

Figure 3.11: Number of Dutch degree-mobile students abroad and percentage share in total enrolment in the Netherlands in 2006-2011 Source: Nuffic, 2014 Figure 3.12: Top 6 destinations of Dutch degree-mobile students in 2006-2011 (in absolute numbers) Source: Nuffic, 2014 The most preferred destination countries of Dutch degree-mobile students are presented in Figure 3.12. In the period of 2006-2011, the number of Dutch students going to United Kingdom, Belgium, and New Zealand increased faster than the number of students choosing for United States and France for their studies. In the same period, a slight decrease in number of students going to Germany can be observed. 21

Outgoing credit mobility Similar to the increasing trend in outbound degree-mobility in the Netherlands, outbound credit mobility has also increased in recent years as shown by Figure 3.13. As mentioned above, outbound credit mobility increased by almost 50% in the Netherlands. On average in Europe, 4.3% of the student population in 2010/2011 was participating in the Erasmus scheme (European Commission, 2013). This percentage was higher in the Netherlands (6.2%). For comparison, the share of outbound credit mobility was 1.7% in the UK, 5.7% in Germany, and 6.5% in Belgium. Luxembourg has, with 34.2%, the highest incidence of outbound credit mobility. Figure 3.13: Number of outbound Dutch credit-mobile students in 2005-2011 Source: Nuffic, 2014 Dutch university students enrolled in fields such as agriculture and natural environment, and engineering and healthcare are more inclined to participate into credit-mobility programs. For students in universities of applied sciences, outbound student mobility is most common in fields such as economics, language and culture, and agriculture and natural environment (Nuffic, 2012). The main destination countries preferred by Dutch credit-mobile students are presented in Figure 3.14. Spain, United Kingdom and Germany are among the countries that have been increasingly receiving Dutch students in recent years. 22

Figure 3.14: Top 5 destinations of Dutch Erasmus students (100 or more Erasmus students) in 2007-2012 Source: Nuffic, 2014 3.3 Net study migration The data discussed in this chapter show that the Netherlands receives considerable amount of degree-mobile and credit-mobile students, and that these numbers have been increasing. All in all, the share of degree mobile students is particularly high is specific fields of study. Degree mobility among Dutch students has also been increasing in recent years, similar to the trend observed all over the world. However, the figures above reflect that the Netherlands is still a net student-receiving country despite the increasing outbound mobility among Dutch students. In 2010/2011, the latest available year with data on both incoming and outgoing degree mobility, more than 52,000 students from outside the Netherlands were enrolled in Dutch higher education. In that year, about 20,600 Dutch students were studying abroad. 23

4 Sensitivity of labour markets forecasts to study migration 4.1 Inflow of graduates and degree migration With its labour market forecasts model, ROA quantifies the expected demand and supply by education and occupation over a six year-period. The confrontation of future demand and supply is used to derive an indicator for the discrepancy between the two (ROA, 2014). This discrepancy indicator the Indicator Toekomstige Arbeidsmarktperspectieven (ITA) can be viewed as an early warning imbalances between demand and supply on the labour market. The ITA is the ratio of expected supply to expected demand in 6 years from now: ITA e,t+6 = L e,t+u e,t +I e,t:t+6 L e,t +O e,t:t+6 +S e,t:t+6, where e stands for various types of education defined by level and field of study, L e,t is the number of people employed in t for education type e, U e,t is the number of people who are short-term unemployed in t for education type e, I e,t:t+6 is the expected inflow of graduates from education type e in the next 6 years, O e,t:t+6 is the expected number of job openings for people with education type e in the next 6 years (it is the sum of the replacement demand and the positive expansion demand, see ROA, 2014), and S e,t:t+6 is the expected substitution demand for people with education type e in the next 6 years. 5 It is obvious that both supply and demand by type of education determine the perspectives for a particular type of education e. When the expected supply exceed the expected demand, i.e. when ITA > 1, then the labour market prospects are bad. This does not mean that young graduates by definition will face unemployment. There are other mechanisms for adjusting to such imbalance between supply and demand. For example, young graduates could seek employment in jobs below their education level, in other fields, accept lower wage offers or jobs with few secondary benefits, or even migrate abroad to seek employment. In the latest labour market forecasts by ROA (2013) it is concluded that the prospects for young graduates from university and university of applied sciences are, on average, poor. Although these prospects differ strongly by field of study, the driving factors behind these poor prospects are 1) the absence of job growth due to the lasting economic crisis, 2) the low replacement due to the increased labour market 5 The ITA can also be calculated from percentages or proportions by dividing all the terms by the current labour supply L e,t : ITA e,t+6 = 1+U e,t Le,t + I e,t:t+6 Le,t 1+ O e,t:t+6 Le,t + S e,t:t+6 Le,t. This shows that the expected inflow of graduates is an important component of the future labour supply (about 88% of the total expected supply). 24

participation at older ages, and 3) the high expected inflow of graduates possibly because of delayed entry in the labour market (Fouarge and Meng, 2014). ROA s labour market inflow forecasts by type of education are derived from forecasts by the Ministry of education (Ministerie van Onderwijs, Cultuur en Wetenschap, 2013). These forecasts include foreign students enrolled at Dutch higher education institutions. One could therefore wonder to what extent the forecasts are sensitive to inclusion of these degree migrants. This is especially relevant in the light of the rising internationalisation in higher education, and the large differences in the incidence of degree migration across fields of study. Rational for including degree migration in inflow forecasts The implicit assumption in calculating the expected inflow (I e,t:t+6 ) is that, ex ante, all degree migrants are here to stay. It should be noted, however, that including degree migration in the inflow forecasts makes much sense from the point of view of the potential contribution of foreign students to the Dutch labour market as discussed in Section 2. For employers, study migrants constitute a pool of potential workers from which they can hire those with the skills required to fill their vacancies. From Dutch graduate s point of view this means they potentially have to compete with degree migrants for the jobs on the Dutch labour market. If labour is scarce, employers will make interesting job offers in order to attract recent graduates, including degree migrants. In case of oversupply on the Dutch labour market, one could expect degree migrants to seek employment elsewhere, but also Dutch students could adopt this strategy in order to improve their employment prospects. Ex ante, degree migrants impact on the expected labour market prospects of natives. Inflow corrected for degree migration Nevertheless, one can test the sensitivity of the labour market forecasts by accounting for differences in the incidence of degree migration across fields of study as reported in Section 3. For this purpose, we computed alternative measures for the inflow (I e,t:t+6 ), where we discount the forecasted inflow for the share of degree migrants by field of study, and the expected retention of degree migrants after completion of their study: I e,t:t+6 = I e,t:t+6 (I e,t:t+6 m e,t (1 r e,t )), where m e,t stands for the share of study migrants in field of study e at time t, and r is the retention rate. If there are no migrants in a specific field of study (m e,t = 0) or if = the all degree migrants stay on the Dutch labour market (r e,t = 1), then I e,t:t+6 I e,t:t+6. In other cases, I e,t:t+6 < I e,t:t+6, and the overestimation of the inflow will depend on both the incidence of degree migration and the retention rate of migrants. 6 6 Note that the inflow could be corrected upward for Dutch students who study abroad since they are not included in the forecasts by the Ministry of Education. However, we disregard these students because of the low incidence of degree mobility (only 3% of Dutch students go abroad 25

The incidence of degree migration by fields of study is taken from Nuffic (2014). This data is discussed in Section 3, and makes the distinction between degree migration among university graduates and graduates from university of applied sciences. The retention rate is not available by field of study. Instead, we use the estimates reported in Section 3. The lower bound estimate suggests the retention is only 19% (r e,t = 0.19). Other estimates suggest a retention rate of 43% (r e,t = 0.43). In addition we compute the inflow under the assumption that none of the degree migrants supply their labour on the Dutch labour market (r e,t = 0). We compare the forecasted inflow and the ITA under these various assumptions to the ITA from the original forecasts from ROA (2013) that assume that ex ante r e,t = 1. 4.2 Sensitivity of forecasts Table 1 reports on the sensitivity of the labour market forecasts for degree migration among graduates from university of applied sciences and university graduates. The ITA, as reported in ROA (2013), equals 1.12 for university of applied sciences and 1.11 for university graduates, which is indicative of poor labour market prospects. Assuming that all degree mobile graduates from university of applied sciences leave the Netherlands upon graduation would reduce the expected inflow until 2018 by 8.1%. For the inflow from university, this would reduce the total inflow by 13.2%. Although this extreme assumption does reduce the ITA, the labour markets prospects until 2018 would still be poor. Obviously, other assumptions with respect to the retention rate of degree migrants lead to a lower reduction of the inflow (7.3% lower inflow from university), and have no impact on the forecasts. to obtain a degree), and because there are no data to on outgoing degree mobility by field of study. 26

Table 1: Sensitivity of inflow and ITA for university of applied science and university graduates, 2013-2018 Original forecasts 1) University of applied sciences (HBO) University (WO) ITA 1.12 1.11 Characterisation ITA poor poor r e,t = 0 Change in inflow (%) -8.1-13.2 ITA 1.10 1.08 Characterisation ITA poor poor r e,t = 0.19 Change in inflow (%) -6.5-10.7 ITA 1.10 1.09 Characterisation ITA poor poor r e,t = 0.43 Change in inflow (%) -4.4-7.3 ITA 1.11 1.10 Characterisation ITA poor poor 1) ROA (2013: p. 14) As discussed in Section 3, degree migration strongly differs across fields of study, so it is possible that discount for it would affect some fields of study more than others. Table 2 shows for graduates from university of applied sciences that correcting for degree migration has a large impact on the forecasted inflow for studies that attract large numbers of foreign students such as economics, social-cultural sciences, and green studies. However, only for the latter does the correction for degree migration affect the forecasts, but only under the assumption of zero retention. 27

Table 2: Sensitivity of inflow and ITA for graduates from university of applied sciences, 2013-2018 HBO onderwijs HBO sociaalcultureel HBO groen HBO techniek HBO paramedisch HBO economie Original forecasts 1) ITA 1.00 1.21 1.01 1.02 1.10 1.21 Characterisation ITA good bad fair fair poor bad r e,t = 0 Change in inflow (%) -5.6-10.9-7.1-4.8-6.0-9.0 ITA 0.99 1.18 1.00 1.02 1.09 1.19 Characterisation ITA good bad good fair poor bad r e,t = 0.19 Change in inflow (%) -4.5-8.9-5.8-3.9-4.9-7.3 ITA 0.99 1.19 1.00 1.02 1.09 1.19 Characterisation ITA good bad good fair poor bad r e,t = 0.43 Change in inflow (%) -3.1-6.0-3.9-2.6-3.3-4.9 ITA 1.00 1.20 1.00 1.02 1.10 1.20 Characterisation ITA good bad good fair poor bad 1) ROA (2013: 67) For university graduates, where degree migration is high for green studies and economics, correcting for degree mobility reduces the expected inflow, but this has no effect on the characterisation of the ITA (Table 3). 28

Table 3: Sensitivity of inflow and ITA for university graduates, 2013-2018 WO letteren en sociaalcultureel WO groen WO techniek WO medisch WO economie en recht Original forecasts 1) ITA 1.10 1.14 1.03 1.00 1.24 Characterisation ITA poor poor fair good bad r e,t = 0 Change in inflow (%) -11.7-23.6-12.6-4.9-16.2 ITA 1.08 1.07 1.01 0.99 1.19 Characterisation ITA poor poor fair good bad r e,t = 0.19 Change in inflow (%) -9.4-19.1-10.2-4.0-13.1 ITA 1.08 1.08 1.01 0.99 1.2 Characterisation ITA poor poor fair good bad r e,t = 0.43 Change in inflow (%) -6.4-13.0-6.9-2.7-8.9 ITA 1.09 1.10 1.02 0.99 1.21 Characterisation ITA poor poor fair good bad 1) ROA (2013: 67) For some specific fields of study, the correction for degree migration does have a relatively large impact on the inflow. Under the extreme assumption that none of the degree migrants remain on the Dutch labour market, the inflow drops with 23.6% for university graduates in environmental science, while it drops with 21.2% for university graduates in business administration and economics. All in all, even in this extreme case, the perspectives only improve for three fields of study (the average reduction of the ITA equals 0.02). The perspectives for HBO environmental sciences and food technology switch from poor to fair with a reduction of the ITA of only.01 point. The perspectives for HBO (physio)therapy switch from bad to poor with a reduction of the ITA change by.02 point. The perspectives for WO economics/econometrics switch from bad to poor with a reduction of the ITA by.06 point. 29